A lightweight FNO classifier achieves 95% F1 on the STEAD dataset even with sparse training data and 98% F1 on real microseismic data, with substantially lower compute cost than typical deep learning models.
Reservoir geomechanics
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.geo-ph 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Seismic event classification with a lightweight Fourier Neural Operator model
A lightweight FNO classifier achieves 95% F1 on the STEAD dataset even with sparse training data and 98% F1 on real microseismic data, with substantially lower compute cost than typical deep learning models.